Tramacere, I Pelucchi, C Bonifazi, M Bagnardi, V Rota, M Bellocco, R Scotti, L Islami, F Corrao, G Boffetta, P La Vecchia, C Negri, E eng Meta-Analysis Research Support, Non-U.S. Gov't Review England 2012/02/24 06:00 Ann Oncol. 2012 Nov;23(11):2791-8. doi: 10.1093/annonc/mds013. Epub 2012 Feb 22. ; International audience ; BACKGROUND: Whether an association between alcohol drinking and non-Hodgkin lymphoma (NHL) risk exists is an open question. In order to provide quantification of the issue, we carried out a meta-analysis of published data. METHODS: We identified 21 case-control and 8 cohort studies, including a total of 18,759 NHL cases. We derived meta-analytic estimates using random-effects models, taking into account correlation between estimates. RESULTS: The overall relative risk (RR) of NHL for drinkers versus non-drinkers was 0.85 [95% confidence interval (CI) 0.79-0.91]. Compared with non-drinkers, the pooled RRs were 0.88 for light (=1 drink per day), 0.87 for moderate (1 to /=4 drinks per day) alcohol drinking. There was no association for light drinkers in cohort studies, whereas for moderate and heavy drinkers, the RRs were similar in case-control (0.85 for moderate, 0.92 for heavy) and cohort (0.89 for moderate, 0.79 for heavy) studies. The inverse relation with alcohol consumption (drinkers versus non-drinkers) was similar in men (RR = 0.83) and women (RR = 0.86), but apparently stronger in studies from Asia (RR = 0.69) than other world areas (RR = 0.88). CONCLUSION: This meta-analysis provides quantitative evidence of a favourable role of alcohol drinking on NHL risk, though the lack of a biological explanation suggests caution in the interpretation of results.